Analysing wages and labour institutions in China: An unfinished transition

A linear wage transition?

Contrary to many developing countries, China has long-term experience with industrial employees. Under the thumb of Maoism, wage systems responded to the need to mobilise the workforce. The notion of sacrifice in the name of national independence was promoted, and this notion enabled the practice of low monetary wages, compensated by benefits in kind (the ‘iron rice bowl’) that depended on the size, prestige and rank of the work unit (danwei 单位) (Walder, 1986). Wage scales were reduced to an extreme degree, and the employee hierarchy was based essentially on seniority (Korzec and Howe, 1992). This socialist heritage of poorly paid employees then played an important role in the economic development of contemporary China.

Indeed, beginning with reforms starting in 1979, an end to egalitarian principles resulted in the introduction of incentive wage schemes (low base salary, supplementation in kind, and a bonus system relying on overtime as the only way to improve one’s wages). This competitive wage relationship, which quickly developed in the coastal provinces, was also combined with the possibility for companies to make profits, leading to an economic transition that was ultimately unfavourable for workers. Indeed, the form adopted for putting the rural labour force to work was initially that of factory wage labour in Township and Village Enterprises or joint ventures in which the social benefits of the danwei were challenged.

According to official Chinese statistics, the real annual wages of urban workers remained low until the mid-1990s (Figure 1), growing at an average annual rate of only 5.9%, ² compared with the real growth rate of the Chinese economy which reached an average of 9.9% per year for this period. The growth of real wages accelerated in the period that followed: from 1998 to 2010, it averaged 15.4% annually, compared to average GDP (gross domestic product) annual growth rate of 10.8%. This trend continued from 2010 to 2015, real urban wages growing on average by almost 13% per year, while average GDP growth slowed to 7.6% per year. Li et al. (2012) emphasise that Chinese wages in manufacturing industries have increased more quickly than in other emerging Asian countries. Compared to India or Indonesia, the Chinese industrial workforce has quickly caught up in most sectors and for all skill levels.

OPEN IN VIEWER

Figure 1.

Real annual wages of Chinese urban workers 1978–2015 (RMB).

However, the gap between the highly industrialised coastal provinces and inland provinces persists (Figure 2): while wages are improving just as much for export industries as for others, Hering and Poncet (2009) show that the wage dynamic is influenced by proximity to export markets. This uneven wage dynamic has contributed to reinforcing inequalities, China becoming one of the countries where the Gini index is among the highest, rising from 0.26 in 1978 to 0.46 in 2014 (Li, 2016).

OPEN IN VIEWER

Figure 2.

Annual average wages of urban workers by region 2004–2015 (RMB).

Nevertheless, inequalities have been slowly shrinking since 2009, owing to the economic crisis, which had a greater impact on the richest provinces (Zhang 2016b). The strong increase in wages is related to the combination of two long-term phenomena: the quick development of manufacturing industries, and the slowdown in the growth of the working-age population (15-64 years) since the beginning of the 2000s (Das and N’Diaye, 2013).

Wages in China: A Kuznets–Lewis textbook case?

The rise in real wages and in wage inequalities illustrates the theories of Kuznets (1955) and Lewis (1955). According to Kuznets, a rise in income inequalities, especially between cities and the countryside, is a normal step in the industrial transformation of an agricultural society, and is one that accompanies the increase in overall income. The expansion of industrial employment creates income and wage inequalities. This phenomenon is a necessary part of the capital accumulation process, since a greater share of GDP is dedicated to profits. In China, the share of wages in GDP has indeed declined constantly from the end of the 1980s to the mid-2000s, while labour productivity has been growing at an ever-increasing pace (Piovani, 2014).

As a country becomes richer, wages in manufacturing industries, though higher than income drawn from agricultural activity, are maintained at a low level due to the movement of agriculture workers entering factories (Lewis, 1955). In China, the abundant reserve of rural workers sustained a long decline of the wage share from the 1980s on. However, Lewis indicates that over time, the demand for skilled labour increases alongside industrialisation and in the end exceeds the supply. This leads to an increase in qualified workers’ wages, and to rising inequalities within the industry.

As long as the turning point has not been reached – while growth in labour productivity remains higher than wage growth – industrial sectors become increasingly profitable, encouraging capital accumulation. Then, when the reserve of agricultural workers is used up, industrial wages grow quickly to the detriment of profits and investment. Economic growth can then no longer rely on low labour force costs, and must instead rely on industrial upgrading.

The acceleration of wage growth since 2003–2004, combined with the sporadic appearance of rural migrant worker shortages in coastal export provinces, could indicate that the Lewis process has begun (Cai, 2018; Das and N’Diaye, 2013; Zhang et al., 2018), thereby contributing to the convergence of rural and urban wages (Cai and Du, 2011). On the other hand, other studies argue that China is far from having used up its demographic dividend, claiming that shortages of low-skilled workforce are real but localised. Rural residents are more inclined to find employment close to home. Spatial disparities are declining as firms from coastal provinces transfer capital and technology inland, where labour productivity is catching up (Lemoine et al., 2015). As emphasised by Knight et al. (2011), while the minimum wage policy has been generally well implemented since 2004, the response of labour-intensive industries in the coastal provinces has been an increase in the number of hours worked, thereby limiting this policy’s benefits. The development policy for the Western provinces (the Western Development Plan launched in 2001 – 西部大开发) and Central provinces (Rise of Central China Plan, adopted in 2004 – 中部崛起计划) also reinforced the scarcity of low-skilled workers in coastal provinces (Zhu and Pickles, 2014).

According to the Migrant Workers Monitoring Survey Report (National Bureau of Statistics of China (NBS), 2016), local migrations have begun to supersede long-distance migrations since 2010. The trend reversal is sustained by the faster growth pace of local migrations vis-à-vis long-distance ones since 2010: in 2016, the migrant worker population fell by 0.3% in the coastal regions and grew by 5.3% in the western provinces. Rural migrants can more easily find jobs in their locality, thanks to the economic development of inland provinces (Gao et al., 2018).

The current situation is thus more complex than the mere depletion of the available rural workforce. Labour shortages in the most industrialised areas of coastal provinces are combined with a persistent rural workforce surplus (approximately 110m workers) ready to accept very low wages, but no longer attracted by the low benefits of migration (Song, 2017).

These analyses are based, however, on a view of the labour market that ascribes the determining role in the evolution of wages to the relative scarcity of labour. Yet Kuznets emphasised that economic growth cannot benefit employees as long as they are kept in an unfavourable power relationship: only institutional innovations, such as public intervention in favour of a minimum wage or employee collective bargaining, can encourage a move from ‘industrial upgrading’ to ‘social upgrading’, ³ which would contribute to reducing income inequalities. While China fits well into the growth pattern related to demographic transition, the institutions that have kept workers’ bargaining power at a low level have equally played a role.

This situation echoes the works of early Institutional Economists like JR Commons (1934) for whom wage stagnation and rising inequalities are a natural trend for capitalism. The employment relationship is indeed an unbalanced institutional structure which creates a structural inequality of bargaining powers, preventing wage-earners from benefitting from productivity gains (Kaufman, 2018). Consequently, wages are not only regulated by the laws of supply and demand as proponents of the Lewis turning point imagine, but are above all the result of power asymmetries between employers and workers.

Industrial development that does not benefit workers

In China, the increase in real wages is constantly outpaced by labour productivity gains: Li et al. (2012) argue that the growth of labour productivity between 1978 and 1997 outpaced the growth of real wages, resulting in a relative decrease in labour costs. This trend has persisted since then (Figure 3). While labour productivity kept on improving, wages lagged behind, especially in the inland provinces manufacturing industries (Cui and Lu, 2018), in relation to industrial relocations.

OPEN IN VIEWER

Figure 3.

Labour productivity and average labour compensation in China, 2000–2013 (100 = 2000 base year).

The intensification of work is reflected in the ongoing drop in the share of labour income in GDP – confirming the weak bargaining power of labour (Figure 4) – at least until the 2008–2009 crisis. A regular rise in minimum wages was then decided by the central government when a minimum wage policy was mobilised to reverse the trend. But the continuation of this policy is now questioned (Li et al., 2018). In China, the level of minimum wages, usually the bottom-line in labour-intensive industries such as garment and electronics, is set locally. ⁴ It is therefore subject to inter-regional competition, as the objectives of attracting investments and industrial development prevail for local governments. This situation leads to a ‘race to the bottom’ in local labour standards and weak law enforcement, endorsed by local authorities (Friedman and Lee, 2010). For instance, the annual increase in mingongs’ (internal migrant workers’) wages has dropped from 21% in 2011 to only 7.4% in 2016 and 6.4% in 2017 (NBS, 2016, 2017).

OPEN IN VIEWER

Figure 4.

China’s Labour Share in GDP (%), 1978–2015.

However, as Xu et al. (2015) demonstrate, workers have always been rewarded less than half of their marginal product, even recently. This is especially true in the private sector, despite the intensive restructuring of the public sector that started in 1992. Labour is indeed subject to capitalist structural constraints, which undermine workers’ bargaining power and prevent labour from getting a larger share of the value added. Consequently, real wages have suffered due to new strategies: increasing the number of working hours for the same wage, the generalisation of a new type of informal labour (temporary workers and interns) after the adoption of the 2007 Labour Contract Law, and relocation of factories to inland provinces where wages are lower. Finally, massive investments promoting automation of labour-intensive industries are making China the largest installer of robots.

The strategy of maintaining low wages is now complemented by the objective of industrial upgrading in Coastal provinces, as illustrated by the ‘Made in China 2025’ plan and massive investments in new technologies (Fabre, 2018). The result has been relocations towards inland provinces and outside China, as the decline in China’s processing trade during the last decade indicates (Chen et al., 2018): the aim is endogenous development and the growth of a domestic market in which MNCs can find their new interest. Therefore, the capitalist logic of keeping wages under the level of their marginal product as depicted by early Institutional Economics can still shed light on the Chinese case (Kaufman, 2018). Despite a large number of jobs created, many have also been destroyed and employees are still subject to an unfavourable power relationship, since Chinese companies and MNCs capture a rent – that is, an income not related to a productive contribution – a well-established characteristic of GVCs.

Industrial specialisation dependent on GVCs

China’s industrial specialisation relies on artificially low wages particularly attractive for Western brands: its economic opening at the beginning of the 1980s coincided with the new offshoring strategies of Western MNCs, in the wake of financialisation and the search for an increased return on equity leading to outsourcing. Since then, the fragmentation of production process within GVCs has been part of a system of economic power relations that gives Chinese firms no other choice but to accept or reject the conditions concerning prices, deadlines and volumes established by Western buyers that own a brand or advanced technology. These ‘lead firms’ have the power to subcontract operations requiring low-skilled labour in low-wage countries (Gimet et al., 2015). ⁵ The pressure on wages is thus non-negotiable for subcontracting factories, and its corollary is a decline in employee bargaining power (Barrientos, 2013; Hayter, 2004). The complex fragmentation of the production system forces workers from different countries to compete on similar tasks (Nathan and Sarkar, 2011).

The mechanism for extracting a rent (i.e. a persistent gap between labour compensation and productivity due to weak countervailing institutional corrections (Commons, 1909, 1934) is based on the ability of lead firms to impose their conditions to suppliers and manage the contractual relation only through authority. The system is indeed globally organised to reinforce the power of lead firms over subcontractors: while the former externalises low value-added tasks, the latter is constantly under the threat of losing orders if its competitiveness deteriorates (Barrientos et al., 2011b; Gereffi et al., 2005).

In this unbalanced contractual relationship, conditions of employment and wages become residual variables, as JR Commons understood a century ago. The subcontractor may potentially increase its margins by adopting a more efficient production process, but only until its competitors adjust their own technology. It is thus almost impossible to establish a ‘sustainable price’ which would allow suppliers to pay decent wages and comply with labour law: globalised capitalism always shifts unreasonable costs onto workers, at the price of the workers’ economic insecurity (Commons, 1909).

The prospect of industrial upgrading, through the development of services and/or products with higher value-added and specialisation in more sophisticated tasks, is often seen as an opportunity to improve workers’ lot (Zhu and Pickles, 2014). Based on this objective, policies to increase the cost of labour adopted by the central Chinese authorities – through minimum wage policies and more stringent labour legislation – become tenable. Yet concretely, the beneficial effects of industrial upgrading on workers’ lives have for the time being been held back by industrial strategies that generally remain fixed on minimising labour costs. The development of more sophisticated technological production has not resulted in the development of internal labour markets within enterprises or in the improvement of the employment relationship for skilled workers. Rather, strategies involving automation, reconfiguration of production systems, or lean management are leading to a reduction in the labour force and a refusal to improve wages and working conditions, or to relocations to South-East Asian countries. The production model based on exploitation of a low-cost labour force is thus very resilient (Butollo, 2015; Lüthje and Butollo, 2017). In that sense Almeida (2009) shows that China is the only example in East Asia’s middle-income countries where the openness to foreign investments and the adoption of recent technologies did not increase the demand for skilled labour. Stronger integration in international markets is associated with a greater specialisation in relatively low-skilled intensive productions, similar to low-income countries like Vietnam and Cambodia.

Even if the Chinese manufacturing system has been able to create more value-added goods for exports since the mid-2000s, these improvements remain limited, even in the tech-intensive industry. Within GVCs, China has so far not been able to attract higher value-added productions (Wen, 2018), and even in IT and electronics, remains specialised in tasks relying on low-skilled and low-paid workforce in most of its manufacturing industries. Despite important productivity gains, Chinese firms’ integration into GVCs did not materialise in improved wages or quality of employment. Compensation conditions continue to depend both on the local labour market and the requirements of foreign buyers.

The GVC analysis reveals that in addition to controlling the segments creating most of the value, lead firms can impose their control on the distribution of that value at each step along the chain. Gimet et al. (2015) indicate that the fragmentation of production leads to a decline in wages over the long term, even more so when MNCs’ strategies force producers at the bottom of the chain to specialise in narrower segments, which hampers their ability to upgrade their production and increase their share of value-added. In other words, contrary to what was being heralded by economic growth theory, integration of Chinese firms in GCVs does not lead to social progress as long as it is based on MNCs organising competition within their suppliers’ network, which leads to a race to the bottom.

The return of the sweating system

This result has already been analysed by JR Commons (1909), in his article ‘American Shoemakers’ as the logical and inevitable product of a ‘competitive menace’. The employer’s greater bargaining power on the labour market, which leads to an imbalance of bargaining power within the market (Kaufman, 2010: 437), has every chance of being reflected in an imperfectly competitive wage. This would be less than a competitive wage, especially since the labour supply is not flexible (i.e. workers need to work and do not withdraw from the labour market) (Kaufman, 2010). The ‘sweating system’ refers to a situation in which labour compensation is aligned with what the marginal worker is forced to accept (possibly a subcontracted worker and/or a migrant worker). This process by which the worker is no longer able to earn wages that match his or her productivity results from changes in the structures of production organised according to the dissociation of the wage from the series of prices negotiated between the various market intermediaries on a large scale. When MNCs decided to externalise their production by making use of an increasingly large set of suppliers, they simply deployed this sweating system on a global scale (McIntyre and Ramstad, 2004).

In the analysis proposed by Commons (1909), the decision to use subcontractors rather than direct hiring – that is, the preference for subcontracted work in small workshops specialised in one segment of the production process, and even for home workers – comes from market expansion. The direct relationship of the client with the producer, which ensures the wage/sales price link, is replaced by a series of intermediaries, which introduces a series of market transactions that separates the determination of the sales price for goods from the wage of the person that produced them. JR Commons contends that the control of information (especially concerning the difference between regular and overtime hours, and therefore the exact compensation to be paid) by the ‘merchant capitalists’ (the MNCs or their contractors) throughout the process is essential for this system to develop, while allowing the capitalists to make a profit without hiring a single person. This evolution did not escape Marx (2015 [1887]), for whom the sweating system was related to piece-work pay. He added that ‘the exploitation of the labourer by capital is here effected through the exploitation of the labourer by the labourer’ (p. 391): the foreman responsible for bargaining the price with the buyer holds the responsibility to set the wages.

This dissociation between wage bargaining and price is reflected in the disconnection between wages and labour productivity, thus allowing growth in wages to be kept under the level of growth in productivity. This series of bargaining between intermediaries creates a ‘transactional gap’ between workers and the entity that captures the results of their labour force (McIntyre and Ramstad, 2004). Within GVCs, this transactional gap is even greater since the lead firms do not own the factories and therefore do not have a direct employment relationship with the workers responsible for the production tasks. In the cascade of subcontractors and ‘subletting labour’ (as designated by Marx, 2015 [1887]) that characterises China’s integration into GVCs, competitive menace has become the norm of employment, at least for workers in export industries: they are in general not aware of the legislation on overtime work, nor of most of the existing labour laws.

A situation of competitive menace, the race to the bottom emerges when the labour market becomes entirely dependent on contractual relations (not competitive relations; Nathan and Sarkar, 2011) that are formed outside the market, in transactions between the various suppliers along the value chain. This leads Lakhani et al. (2013) to consider the coexistence within GVCs of several employment relationship configurations that result from buyer/subcontractor relations going from the market to the chain of command depending on the degree of complexity of the tasks to accomplish, the possibility that they can be codified and the capacities of subcontractors: it is the interconnections within and between the firms involved in a GVC that configure the employment relationships in each segment.

It is thus not because the worker does not enter into a direct employment relationship with the lead firm that he or she can escape from the specific aspects of the labour contract; instead, the subordinate relationship here is lost in the series of commercial transactions conducted at a global scale; it is especially the case given that his or her direct employer must be able to ensure the flexibility required by the buyers. This leads to the cohabitation of two types of workforce: one is permanent, legally employed and benefits from social legislation; the other is casual, temporary and often not declared, which always provides the required flexibility (Barrientos et al., 2011a). One of the characteristics of GVCs is thus to create or strengthen the segmentation of the labour market. This trend is confirmed by the split between industrial and social upgrading in the Chinese industry (Butollo, 2015).

This phenomenon of unfair competition is based not only on the fact that the employee bargains without power on the market, but also that it is combined with an imbalance of bargaining power ‘before the market’ (Kaufman, 2010: 437) due to the unequal distribution of property rights that benefit the employer. Workers find themselves in a situation of ‘institutional exploitation’ in which the pay discrepancy between capital and labour reaches a level no longer considered as ‘reasonable’ according to Commons’ definition: the difference between a wage and the minimum job conditions considered by society as reasonable – that is, for which a consensus is negotiated at a specific time – and the real level of wages that results from biased rules of the competitive game. Yet the possibility for institutional exploitation is made easier due to the globalised (transactional gap) and fragmented (strengthening the imbalance of powers) nature of production networks (Rossi et al., 2014). China currently fits in such an ‘unreasonable’ situation of asymmetrical powers due to the persistence of specific labour institutions endorsed by the state.

Labour institutions favouring integration within GVCs

Within GVCs, the employment relationship is thus no longer modelled solely according to a labour contract between an employer and his or her employees but is also based on inter-firm arrangements at a transnational scale (Lakhani et al., 2013). The prevalent labour institutions in China have shown themselves to be highly compatible with lead firms’ strategies. Following JR Commons (1934), institutions are the formal or informal, artificial (i.e. human-made) rules that govern the economic activities. Under the capitalist system, institutions enable accumulation and transactions, and here more specifically, the labour contract.

Labour law that legitimises institutional discrimination against rural migrants

The Chinese labour law developed as part of the reforms to the socialist system was not designed to apply to migrant workers (mingong 民工). Indeed, the household registration system (hukou) contributes to providing subcontractors of lead firms with an opportunity to access a labour force at a much lower cost, as these workers are deprived of basic social and civil rights in the cities. Even if the original hukou implemented in the 1950s was not designed to this end – but rather to prevent rural migrations towards cities – its persistence during the transition process to capitalism has been a major driver in the fragmentation of the Chinese workforce (Chen, 2019). This case illustrates Commons’ theory of institutional evolutionism. Often compared to a ‘floating population’, migrant workers make up the vast reservoir of a workforce that can be exploited at will in workshops or on construction sites (Solinger, 1999). Employers can thus benefit from the semi-clandestine situation of the mingong and employ them without labour contracts, with the active support of local authorities and the official trade-unions, which were in charge with supplying workforce to the joint ventures in the 1980s. This role further benefitted the nascent private sector all along the reforms process.

To be more specific, the introduction of the labour contract in 1986 led to the cohabitation of two types of contract: the hetong zhigong (合同职工), which was intended for employees of public companies (zhigong 职工) who later benefitted from protections in the labour law adopted in 1994, and the old temporary contract (hetong gong 合同工), which was intended for the mingong, and which provided almost no guarantees. The dual nature of the labour market explicitly placed the mingong into an informal wage relationship. This arrangement was a windfall for employers in the coastal provinces. It allowed them, in complicity with local authorities (including the official trade-unions local branches), to adopt a varied implementation of the labour law. They were thus able to enjoy the full benefit of an available and willing workforce, in order to create a competitive advantage that other countries, including in Southeast Asia, have had trouble sustaining.

New forms of labour fragmentation supported by local governments

At the national level, labour legislation has been strengthened based on a well-calculated strategy to raise the cost of labour. These policies aimed at moving migrant workers into an economically secure status (Périsse, 2017) and at fostering industrial upgrading in inland and western provinces. ⁶ However, companies integrated within GVCs find in local governments an ally to circumvent these regulations and to maintain low production costs, as illustrated by the new production complexes developed by Foxconn in the last decade in central and western provinces.

The Taiwanese company Foxconn Technology became, in the 2000s, the world’s main supplier of electronic devices, now assembling smartphones and laptops for the largest brands, such as Apple and Samsung. It developed in the city of Shenzhen (Guangdong province) a specific production model relying on massive use of low-cost labour, and a ‘neo-Taylorist’ work organisation (Lüthje and Butollo, 2017). Facing rising labour costs and the development of strikes in the wake of the 2008–2009 crisis, Foxconn started negotiating with several other municipalities to establish new production sites. Local governments in the cities of Zhengzhou (Henan) and Chengdu (Sichuan) provided numerous benefits to the company, such as free recruitment offices, transportation services to the factories, cuts in corporate and value-added taxes, and partial funding of the manufacturing complexes (Barboza, 2016; Chan et al., 2015).

More fundamentally, the Sichuan and Henan provincial governments supported and organised the large-scale recruitment of interns in vocational schools. Top-level officials now assign targets that lower administrative levels must complete to ensure that thousands of student workers are brought on the production lines (Chan et al., 2015). While such preferential treatment by local governments to capital is not new, it has thus supported, in the last few years, the development of a new form of labour fragmentation.

In China, as elsewhere, globalised production is driving the development of new forms of subcontracted bonded labour (Barrientos, 2013; Lüthje, 2014a). The production model based on low labour costs is difficult to overcome, and the Chinese labour institutions enable the sweating system to become more sophisticated, thus increasing economic insecurity for non-permanent workers (Lüthje, 2014b; Périsse, 2017): the mingongs’ labour contract coverage has dropped since the 2008–2009 crisis, and in 2016 sat at 35.1%. The fragmented Chinese labour market can therefore be understood as a systemic combination of the GVCs’ logic and the specific rules driving the labour force to industrial work, with a strong role for the state in defining those rules.

A ‘Chinese-style’ industrial relations system

The inability to improve workers’ compensation at the same pace as productivity is a source of social conflict. A recent study has found that wealth distribution has become the main concern of Chinese workers, as more than 80% of collective action between 2015 and 2017 demanded the payment of wages in arrears and wage increases (China Labour Bulletin, 2018). However, the ambivalent position of the All-China Federation of Trade Unions (ACFTU) prevents the emergence of explicit regulations on wages involving workers’ representatives. Preoccupied with controlling the negative effects of accelerated industrialisation, the Chinese authorities tightened their grip on the official trade union after the Tiananmen Square massacre in 1989. Besides controlling workers, the ACFTU was explicitly granted by the Communist Party the role of maintaining ‘social stability’ (Clarke and Pringle, 2009). Since then, the ACFTU has become a quasi-governmental organisation in charge of representing workers within the state (i.e. to provide feedback on labour problems from the grassroots as a base for new labour regulations) and to manage labour conflicts swiftly. Its action is twofold.

Instead of the ACFTU being a counterweight to employers, workers are prohibited by the law from defending their own interests: the trade union seeks to maintain competitive wages, a major concern for municipalities responsible for local development. On the other hand, the ACFTU must limit any social troubles by incorporating the new working class into urban society, especially by making migrant workers theoretically eligible for labour rights. The power imbalance in the wage relationship nonetheless persists due to the inability of the trade union to truly defend the rights of the mingong within factories. According to Song (2017), the presence of the ACFTU in companies does not have a positive effect on wages. ⁷ The lack of legitimate workers’ representatives contributes to the high level of conflicts in China. The ACFTU may provide individual legal assistance to workers, but never helps them to organise collective action, thus indirectly fostering spontaneous and often violent labour conflicts. Here is another example of institutional evolutionism: as an official part of the local Communist Party organisations, the ACFTU contributes to the economic goals of local authorities, which typically implies siding with employers in case of labour conflict and preventing any emergence of a trade-union alternative. The ACFTU is indeed given by the law the monopoly of representation of the Chinese workers (Clarke and Pringle, 2009).

It is therefore in the field of conflict management institutions that the official Chinese trade union has been creative, as demonstrated by its ‘cuts and pastes’ from labour relations institutions advocated by the ILO (Lee, 2009). For the government, these so-called social dialogue institutions may somehow contain conflict escalation; however, employers who retain exclusive rights in defining employment conditions see no point in genuine negotiations. The wage policy is viewed as being the sole privilege of the employer and is often kept secret so that employees cannot grasp the mechanisms of wage negotiations (Chan and Siu, 2010). A lack of negotiations on social benefits is also widespread, and there is an almost total lack of seniority-based promotion, and of job classification system within factories (Lüthje et al., 2013). On the other hand, entrepreneurs are suffering from this restrictive view of labour relations, which fosters experienced and skilled workers’ turnover and accentuates competition among firms to retain them. These labour problems are considered by Commons as a natural situation as long as the labour relation is only managed through authority. The process of collective bargaining therefore appears as the only way to avoid destructive conflicts in the capitalist system – what he calls ‘reasonable capitalism’. However, in China, the question of autonomous representation of workers remains out of sight despite rising discontent. ⁸

The rise in strikes since the 2008–2009 crisis has so far only produced limited new regulation in the matter of social dialogue (Hui and Chan, 2016). Consequently, even the wages of skilled urban workers have stagnated: the 0.9% rise in wages for white-collar workers in 2017 only benefits a minority of ‘senior managers’, while inequalities are increasing between workers in big cities and ‘clerical workers’ in medium cities (China Labour Bulletin, 2017). Overall, Chinese labour institutions ensure the maximum individualisation of the employment relationship, thus upholding of a competitive menace on workers by lead firms and their Chinese suppliers (Kaufman, 2010).

Low-wage trap versus Chinese-style Fordism?

The 2008–2009 crisis sounded the alarm and accelerated the transition towards a new economic policy to reorient the growth regime. Since then, Chinese leaders have been seeking to move away from being a low-cost supplier in order to benefit from a better educated and healthier workforce, increase the value-added of production, re-balance distribution in favour of wages and develop domestic demand in order to no longer have to depend on fluctuations in global demand (Zhang, 2016a). This goal still seems very distant as indicated by the decrease in the share of consumption in GDP, which despite a slow rise since 2007 (from 35% to 39% in 2016) remains one of the lowest in the world. However, the share of wages in household income is high in China, the highest in emerging or developing countries (ILO, 2014).

The recovery plan of 2010 has also generated contradictions: it served to develop infrastructure, particularly in underfunded regions, which improved opening-up. But it relied on an increase in investments rather than in household consumption and fuelled overcapacity in heavy industry. Furthermore, local authorities still have the ability to dilute national guidelines by prioritising ‘visible’ and effective economic results in the short term, which has led to a trend to limit as much as possible the effects on wages of pro-labour regulations, especially the 2007 Labour Contract Law.

Chinese growth has therefore relied on companies (all categories combined) extracting high levels of productivity for low wages even during the most recent period (Piovani, 2014; Xu et al., 2015). In the long run, this type of growth regime may turn against accumulation itself. Indeed, the effectiveness of this ‘competitive’ type of capitalism (Boyer, 2011) is ultimately reflected in over-investment: while low wages create insufficient opportunities to develop domestic demand, the profitability of the capital invested is in constant decline (in 2017, investments still represent 45% of GDP). The tacit compromise between Chinese workers and the Communist Party could in the end not be profitable for anyone, as the competition between municipalities to attract investors and the exclusion of workers from the industrial relations institutions combine to hold back workers’ bargaining power (Séhier, 2017a). The prospect of sustainable restitution of productivity gains over the long term through increased real wages as implemented in the form of a post-war Fordist compromise in Western countries seems, so far, out of reach in today’s China.

Impediments to the construction of a Chinese-style Fordist wage relationship

The absence of linkage between productivity and wages has been a characteristic of the Chinese wage relationship for decades (Piovani, 2014; Xu et al., 2015), thus undermining the scenario according to which there could be an evolution towards a Fordist type of wage relationship. The reasons have to be sought in the extreme diversity of production regimes in the key industrial sectors of the Chinese economy. Lüthje et al. (2013) bring to light the great heterogeneity of wage relationships in various industries (automotive, electronics, textiles and clothing), which reflects a large variety of sectoral and regional capital accumulation regimes. So far, insufficient industrial upgrading in the most profitable sectors like electronics thus reinforces the fragmentation of the Chinese production system (Wen, 2018), and the evolution towards a Fordist regime seems unlikely in the near future.

The combination of ‘intensive’ (such as high-tech parks) and ‘extensive’ (low value-added production) capital accumulation strategies within complex production networks is reflected in the coexistence of industrial sectors at different levels of development. These industries are themselves spread throughout regions with various specific features that are maintained by local government actions and combined with the strong divide between urban and rural areas. China could more precisely be described as having a ‘variety of capitalisms within one country’ (Lüthje, 2014b: 6) where the divides inherited from the past have built resilient fault lines that show no overall trend towards an accumulation regime that connects a rise in capitalistic intensity, rapid productivity gains and higher wages.

In every key sector of the Chinese economy, labour standards and job flexibility within enterprises have been institutionalised to varying degrees, based on various combinations of specifically Chinese features inherited from socialism together with influences from Japanese, Western and Taiwanese models. This accounts for the consistent separation both between labour productivity and wages, and between the fast and ongoing enrichment of the country and the constant drop in the share of wages in the value-added. The attempt to reorient the Chinese economy towards the domestic market is beginning to somewhat bear fruit (Lemoine et al., 2015), but this only marginally benefits employees.

This is a situation that JR Commons (1909) considered: an imbalance of powers in the wage relationship that contributes to creating divides within the labour force, which also helps us understand why the attempt of industrial upgrading following the crisis of 2008–2009 did not result in a major overhaul of production regimes and labour practices (Lüthje, 2014b). For the moment, a Fordist accumulation regime, characterised by the redistribution of productivity gains to benefit real wages, does not represent a likely evolution for the Chinese economy, which continues to be characterised by the under-consumption of entire sections of the working class, including the most stable industrial urban labourers. The counterpart is very high household savings which until now has been justified by the almost complete absence of a social welfare system for a large portion of the population, as well as very high education costs.

The dismantling of social insurance provisions has in fact been one of the consequences of the privatisation and reorganisation of the public sector, leading to the exclusion of large segments of the Chinese population (Ngok and Huang, 2014). Efforts have been made to unify the social welfare system with the intention of including the entire working population in health and retirement insurance systems by 2020. The goal of the central government, however, is being met with resistance by local governments responsible for the financial management of this system. Concerned with maintaining their attractiveness to industrial manufacturers, they often baulk at claiming what they are owed from the companies, while the richest provinces delay redistribution to the poorest provinces. Here again, migrant workers are the worst off as the majority among them remain concretely excluded from the social welfare system (Jiang et al., 2018; Séhier, 2017b).

The inclusion of social criteria in the evaluation of local officials represents a small step to foster the implementation of the recent regulations, such as the 2011 Social Insurance Law. However, despite significant progress in terms of the amount of spending and the coverage rate at the national level, the level of insurance granted to each worker remains low. The efforts made to develop indirect wages thus do not compensate for the low increase in direct wages.

The recent escalation in the number of conflicts related to social welfare has shown that collective pressure exerted by the base could require executives and employers to effectively implement reforms that are favourable to workers. For example, the strikes that took place at the Yue Yuen factory in Dongguan in April 2014 united 40,000 workers, while the one at the Lide factory in Guangzhou a few months later led more than 2,500 workers to organise over several months to recover social security back pay. It should therefore not be overlooked that in spite of the hardening of the regime with regard to activists, Chinese officials and the trade union may be forced to accept a greater participation of workers in the development and implementation of social welfare provisions, which is the only way to respond to employees’ need for economic security (Périsse, 2017).

Conclusion

While China seems to have been generally successful in its transition from a planned economy to a market economy, a hasty assessment might lead one to believe that it is an example of a model of development based on Kuznets and Lewis, with the increase in real wages since the end of the 1990s indicating the end of the demographic dividend. A closer examination, however, leads to a more nuanced understanding of this scenario based solely on the mechanics of labour market flows: while Chinese growth since the early 1980s can indeed be characterised by turning rural population into wage labour, it was accompanied by a weak institutionalisation of employment relationships and led to a structural imbalance unfavourable to workers.

This specific institutional configuration has been successful for long-term integration into GVCs. This productive system operates at a global scale under the thumb of large companies headquartered in developed countries, where, due to the contractual nature of the relations between buyers and subcontractors, it is possible to obtain high productivity from Chinese workers, for which wages remain low. Thus, contrary to appearances, even though real wages are rising, the share of wages in the nation’s wealth has not kept pace. This situation in which workers are not able to obtain compensation that is at least equal to their productivity reflects a form of organisation that favours the capitalist entrepreneur, which institutional economist JR Commons (1909) calls a ‘competitive menace’: the employer’s market power leads to non-competitive wages.

Beyond the perverse effects of a development model aligned with capital accumulation (rise in income inequality and rampant social unrest), any evolution towards a Fordist type of accumulation regime remains unrealistic. Even though the government has called for change since the 2008–2009 crisis, Chinese productive structures continue to be affected by the fragmentations inherited from socialism, the various foreign models imported, and the form of integration within global production networks. None of the key sectors of the Chinese economy presents a trajectory towards ’Chinese-style’ Fordism, nor have we seen a shift of productivity gains towards employees. On the contrary, the move towards flexible labour and production imposed by the MNCs and their subcontractors has accentuated wage inequalities within each industry.

This situation is the combined result of an experimental laissez-faire approach during the 1990–2000s, when China’s integration in GVCs benefitted from specific labour institutions, and of the persistent will of the Communist Party to restructure the public sector to erect national champions in strategic fields. Commons (1934) considered that in the building of the legal rules of capitalism, the most powerful social groups are able to influence the state, particularly to protect property rights against labour and social rights. These rules are stabilised until new social groups claim a fairer recognition of their contribution to wealth creation, as is currently the case with the rise in mingongs’ strikes, and their demand of genuine citizenship (Périsse, 2017). Thus, Chinese authorities find themselves faced with a dilemma: how can they foster industrial upgrading and at the same time ensure active economic growth that guarantees employment for all workers, which is the only way to ensure the much-desired social stability?